Bottles of water vary in price with some priced as if they were bottles of fine wine. This article attempts to explain price differences between over 100 bottled waters included in a guidebook to fine waters by drawing on the hedonic pricing approach, which has been used to try to explain price differences among bottles of wine. As part of that approach, the price of each bottled water is regressed against various characteristics, including those related to its water. Water-related characteristics explain only a small part of the price differences among the bottled waters. Thus, to a large extent, the premium that consumers pay for a more expensive bottled water does not seem to be a premium for its water. (JEL Classifications: C21, Q25)

Advanced machine learning techniques like Gaussian process regression and multi-task learning are novel in the area of wine price prediction; previous research in this area being restricted to parametric linear regression models when predicting wine prices. Using historical price data of the 100 wines in the Liv-Ex 100 index, the main contributions of this paper to the field are, firstly, a clustering of the wines into two distinct clusters based on autocorrelation. Secondly, an implementation of Gaussian process regression on these wines with predictive accuracy surpassing both the trivial and simple ARMA and GARCH time series prediction benchmarks. Lastly, an implementation of an algorithm which performs multi-task feature learning with kernels on the wine returns as an extension to our optimal Gaussian process regression model. Using the optimal covariance kernel from Gaussian process regression, we achieve predictive results which are comparable to that of Gaussian process regression. Altogether, our research suggests that there is potential in using advanced machine learning techniques in wine price prediction. (JEL Classifications: C6, G12)

Rank preference and mixture models have been employed to evaluate the ranks assigned by consumers in taste tests of beans, cheese, crackers, salad dressings, soft drinks, sushi, animal feed, and wine. In many wine tastings, including the famous 1976 Judgment of Paris and the 2012 Judgment of Princeton, judges assign scores rather than ranks, and those scores often include ties. This article advances the application of ranking and mixture models to wine-tasting results by modifying the established use of a Plackett-Luce rank preference model to accommodate scores and ties. The modified model is tested and then employed to evaluate the Paris and Princeton wine-tasting results. Test results show that the mixture model is an accurate predictor of observed rank densities. Results for Paris and Princeton show that the group preference orders implied by the mixture model are highly correlated with the orders implied by widely employed rank-sum methods. However, the mixture model satisfies choice axioms that rank-sum methods do not, it yields an estimate of the proportion of scores that appear to be assigned randomly, and it also yields a preference order based on nonrandom preferences that tasters appear to hold in common. (JEL Classifications: A10, C00, C10, C12, D12)

The price of wine is a key topic among market participants interested in valuing their stock, including dealers and restaurants, and consumers who may be interested in optimizing their purchases. A closely related issue, revaluation is the need to regularly update the value of a stock. This need is especially acute in the growing industry of wine as an investment. In this case, fair-value measurement is compulsory by law. We briefly review methods available to funds and introduce a new quantitative method aimed at achieving compliance with IFRS (International Financial Reporting Standard) 13 for fair valuation. Using auction data on 26,640 lots, we apply this method to compute the current fair value of a basket of 232 different wines. (JEL Classifications: C14, C43, Z11)

We present a simple static way of optimizing the prices of bottles of wine for restaurants with a given cellar. In contrast to classical assortment pricing models, we posit that the cellar (i.e., inventory) is given and is not taken as a variable entering the optimization program. In our model, the optimal price is driven mainly by a rating parameter after the effect of initial cost is removed. This parameter plays the role of a dominant characteristic in hedonic models, even though the levels of stocks may also be determinant when they are very low. We provide a numerical sensitivity analysis of prices to various parameters and study a realistic large-scale example based on two wine lists with 50 bottles each. Finally, several extensions are discussed. (JEL Classifications: C61, L11, L83)